[llvm] [LV] Enable considering higher VFs when data extend ops are present i… (PR #137593)

[David Sherwood via llvm-commits]

david-arm wrote:

> > I think the correct place to do this would be in shouldMaximizeVectorBandwidth, since not all targets want to maximise the bandwidth.
> 
> ok thanks @david-arm . Will try to make this target specific and make amendment to AArch64 `shouldMaximizeVectorBandwidth` to start with

I think @SamTebbs33 is working on improving register pressure calculations for partial reductions. It looks like in general with a better cost model we will maximise the vector bandwidth automatically because the phi nodes for the larger types disappear.

https://github.com/llvm/llvm-project/pull/137593